Sex Gender differences in trends in hospital admissions for major cardiovascular events and procedures in people with and without diabetes in England: a nationwide study 2004 – 2014

Running title: Sex Gender differences in cardiovascular events in diabetes

Anthony A Laverty PhD1, Alex Bottle MSc PhD1, Sung-Hee Kim MB BS1, Bhakti Visani

MB BS1, Azeem Majeed MD FRCGP1, Christopher Millett PhD FFPH1, Eszter P Vamos MD

PhD1

1 Public Health Policy Evaluation Unit, School of Public Health, Imperial College London,

London, UK

Email addresses of authors (in order of authorship)

a.laverty@imperial.ac.uk; robert.bottle@imperial.ac.uk; sung-hee.kim08@imperial.ac.uk; bhakti.visani08@imperial.ac.uk; a.majeed@imperial.ac.uk, c.millett@imperial.ac.uk; e.vamos@imperial.ac.uk.

Word count (main text): 4,3193,325. Abstract: 284345.

Tables: 3. Figures: 1.

Address for Correspondence:

Eszter P Vamos MD, MPH, MFPH, PhD

Consultant in Public Health & NIHR Academic Clinical FellowSenior Clinical Lecturer in

Public Health

School of Public Health, Imperial College London,

London W6 8RP, UK

Tel: 00 44 (0) 207 594 0817

Fax: 00 44 (0) 207 594 0854

1

Abstract

Background

Secondary prevention of cardiovascular disease (CVD) has improved immensely during the

past decade but controversies persist on cardiovascular benefits among women with diabetes.

We investigated 11-year trends in hospital admission rates for acute myocardial infarction

(AMI), stroke, percutaneous coronary intervention (PCI), and coronary artery bypass graft

(CABG) in people with and without diabetes by sex gender in England.

Methods

We identified all hospital admissions for cardiovascular disease causes among people aged 17

over 18 years and above between 2004 and 2014 in England. We calculated diabetes-specific

and non-diabetes-specific rates for study outcomes by sexgender. To assess temporal

changes, we fitted negative binomial regression models.

Results

Diabetes-related admission rates remained unchanged for AMI (incidence rate ratio (IRR)

0.99 [95% CI 0.98-1.01]), increased for stroke by 2% (1.02 [1.01-1.03]) and PCI by 3% (1.03

[1.01-1.04]) and declined for CABG by 3% (0.97 [0.96-0.98]) annually. Trends did not differ

significantly by diabetes status. Women with diabetes had significantly lower rates of AMI (

IRR 0.46 [ 95% CI 0.40-0.53]) and stroke (0.73 [0.63-0.84]) compared with men with

diabetes. However, sexgender differences in admission rates for AMI rates attenuated in

diabetes compared with the non-diabetic group. While diabetes tripled admission rates for

AMI in men (IRR 3.15 [ 95% CI 2.72-3.64]), it increased it by over 4-fold among women

(4.3027 [3.7880-4.93]). Furthermore, while the presence of diabetes was associated with a 3-

fold increased rates for PCI and 5-fold increased rates for CABG (IRR 3.14 [2.83-3.48] and

2

5.01 [4.59-5.05], respectively) in men, among women diabetes was associated with a 4.4-fold

increased admission rates for PCI and 6.2-fold increased rates for CABG (4.37 [3.93-4.85]

and 6.24 [5.66-6.88], respectively). Proportional changes in rates were similar in both

sexesmen and women for all study outcomes, leaving the relative risk of events admissions

unchanged.

Conclusions

Diabetes still confers a greater increase in risk of hospital admission for AMI in women

relative to men. However, the absolute risk remains higher in men. These results call for

intensified CVD risk factor management among people with diabetes, consideration of

gender-specific treatment targets, and treatment intensity to be aligned with levels of CVD

risk.

Keywords: cardiovascular, diabetes, gender differences, hospital admissions, acute

myocardial infarction, stroke, coronary revascularisation.

3

Background

Although many developed countries have documented substantial reductions in the incidence

of cardiovascular disease (CVD) in diabetes in recent decades, one half of patients with

diabetes die prematurely from a CVD cause, and over a quarter of all hospital admissions for

CVD are diabetes related (1, 2). In the general population, even in high CVD risk groups, the

risk of fatal and non-fatal CVD events is over 20% lower among women, and women develop

CVD events 5-10 years later in life on average than men (3, 4). However, this pattern

changes in diabetes (5). While diabetes has been reported to double the risk of coronary heart

disease among men with diabetes, it triples it among women with diabetes (6). The risk of

death following an acute myocardial infarction (AMI) is substantially higher in women

compared with men with diabetes (7). Furthermore, a pooled analysis estimated a 27% excess

risk of stroke among women with diabetes compared with male counterparts (8). In the

presence of diabetes, major cardiovascular events have been estimated tocan present 20-30

years earlier in women and 15-20 years earlier in men compared with individuals free of

diabetes (9). Hence, among women, diabetes confers a greater risk of CVD compared with

men to the extent that it eliminates, and to some degree reverses the ‘gender protection’

shown in women without diabetes (10).

The underlying mechanisms of the greater adverse impact of diabetes on CVD outcomes

among women are not fully understood. Contributory factors may include sexgender

differences in biological factors, risk factor profile, and management of diabetes and its

complications (5). While women without diabetes have a more favourable CVD risk profile

compared with men, this association alters in diabetes. Women have shown to be subject to

more pronounced changes in a range of traditional and novel cardiovascular risk factors while

transiting from a non-diabetic to a diabetic state (11). As an example, men develop diabetes

4

at lower levels of mean body mass index compared with women (12). Therefore, women

newly diagnosed with diabetes may already have a more adverse risk profile compared to

men. These differences may be augmented by the underutilisation of evidence-based

preventive and therapeutic interventions in women with diabetes that may, at least partly,

follow from the underestimation of patient risk (13, 14).

Although secondary prevention has become more equitable between men and women with

diabetes in England, women are still less likely than men to achieve treatment targets (15).

However, no previous studies have quantified to what extent substantial investments in

secondary prevention translated into changes in major CVD events at a national level in

England during the past decade, and whether both sexesmen and women benefitted equally

from CVD reductions. National clinical guidelines on the management of diabetes and

quality improvement initiatives in primary care do not consider potential excess CVD risk

among women with diabetes (1, 16). The lack of recent epidemiological studies on this topic

may have resulted in the notion that major CVD outcomes are equal in both sexesmen and

women or follow a similar pattern to that seen in the general population. In light of

considerable targeted efforts to improve comprehensive secondary prevention in diabetes,

potential gender inequalities in hard clinical outcomes have major policy implications.

Therefore, to address this gap in knowledge, our nationwide study aimed to assess trends in

hospital admissions for pre-defined CVD outcomes in males and females with and without

diabetes including AMI, stroke, percutaneous coronary intervention (PCI), and coronary

artery bypass grafting (CABG), and associated inpatient mortality between 2004-2005 and

2014-2015 in England. We also assessed changes in the relative risk of study outcomes in

women relative to men with and without diabetes during the study period.

5

6

Methods

Hospital Episode Statistics (HES) is an administrative dataset that covers data on all inpatient

hospital activity and day case admissions to National Health Service (NHS) hospitals in

England, including private patients treated in NHS hospitals. We used HES data between

2004-2005 and 2014-2015 for all NHS hospital trusts in England. For each hospital

admission, we extracted data on patient demographics (age and sexgender), length of hospital

stay (LOS; inpatient days are calculated by subtracting the day of admission from the day of

discharge), inpatient mortality, principal diagnosis on admission and secondary diagnostic

codes (up to 19) using 10th revision of the International Statistical Classification of Diseases

and Related Health Problems (ICD-10) codes. Procedures were identified using Office of

Population Censuses and Surveys’ Classification of Surgical Operations (OPCS4) codes, that

records up to 12 procedures.

We identified cardiovascular complications as the principal diagnosis on admission,

including AMI (ICD-10 I21 and I22) and stroke (ICD-10 I60–I64). Cardiovascular

procedures were identified using procedure codes for PCI (OPCS4 K49, K50, and K75) or

CABG (OPCS4 K40–K46) in any procedure field. Diabetes status was identified based on

type 1 or type 2 diabetes (ICD-10 codes E10 and E 11) in any diagnostic field. Only the first

admission was counted for patients with repeated admissions for the same cause during the

same year. In total, we identified 754,500 hospital admissions for AMI (476,612 in men and

277,888 in women), 871,331 admissions for stroke (421,757 in men and 449,574 in women),

712,853 admissions for PCI (525,927 in men and 186,926 in women) and 233,882

admissions for CABG (184,919 in men and 48,963 in women) during the study period.

7

Diabetes-specific admission rates were calculated by sexgender and three separate age-bands

(17-44 years, 45-64 years and over 65 years) using the total number of patients with diabetes

in each specific sexgender- and age- band as denominator. To calculate diabetes-specific

admission rates, we Diabetes denominator data was obtained used data from the Quality

Management and Analysis System (QMAS) on the number of people with diabetes aged 17

years and above in England for each study year as denominator (17). QMAS is the financial

database for the Quality and Outcomes Framework (QOF), a national pay-for-performance

scheme introduced in 2004 in UK primary care. QOF rewards general practices for delivering

high quality care and aims to standardise improvements in the provision of primary care

services across a range of areas, including the management of chronic conditions. Under

QOF, general practices are financially incentivised to detect and record all diabetes cases

among their practice populations, as practice payments are weighted by disease prevalence.,

andData in QMAS hasincludes diabetes counts for people aged ≥17 years for virtually all

(>99%) general practices in England. As QMAS data do not contain record age and

sexgender-specific information on diabetes. Therefore, we used obtained data on the age and

sex distribution of people with diabetes in England from the Health Survey for England

(HSE) on the age and sex distribution of people with diabetes in England (18). The

prevalence of diabetes by age group and sexgender were missing in HSE datasets in years

2004, 2005, 2007 and 2008. Similar to previous studies, for the missing years, we used HSE

data from 2006 (the mid-term of this 5- year period), allowing us to calculate age-and

sexgender-specific admission rates for each study year (19, 20). Diabetes-specific admission

rates were calculated by gender and three separate age-bands (17-44 years, 45-64 years and

≥65 years) using the total number of patients with diabetes in each gender- and age- band as

denominator.

8

The admission rate of CVD events in people without diabetes for each year was calculated by

deducting the number of people with diabetes from each corresponding age and sexgender

stratum of the resident population, extracted from Office of National Statistics (21).

Therefore, the denominator for people without diabetes only included the number of people

without diabetes.

The admission rates were directly age- and sexgender standardised using the population

structure in the first year of study period (2004-2005) as the reference population. Rates were

expressed per 100,000 people with or without diabetes. Inpatient mortality rates were

calculated for people with and without diabetes.

Group differences between populations in 2004 and 2014 were tested using Chi-square test

for categorical variables and Student’s t test or Wilcoxon rank sum test for continuous

variables, as appropriate. Female-to-male ratio for each study outcome was calculated as the

ratio of the number of women admitted divided by the number of men admitted to hospital.

Due to evidence on overdispersion, we fitted negative binomial regression models separately

for patients with and without diabetes using age, sexgender and study year as independent

variables. Interaction between and sexgender vs. year were tested for all outcomes in both

groups. In a second set of models including the entire study population (people with and

without diabetes), the relative risk of being admitted for CVD events was estimated for

diabetes and sexgender groups using men without diabetes as the reference. From the same

models, we also report incidence rate ratios (IRRs) for women with diabetes using men with

diabetes as reference. In these models, interactions between diabetes status and year were

tested. Statistical analyses were performed using Stata version 14.0.

9

10

Results

In people with diabetes, there was an increase in the absolute number of admissions for all

study outcomes in both men and women between 2004 and 2014 in England. The baseline

characteristics of people admitted for CVD by diabetes status and sexgender are shown in

Table 1. The largest increase was evident for PCI, as the number of admissions more than

doubled in both sexes men and women during the study period. By contrast, in patients

without diabetes, a decrease in the number of patients admitted for AMI and CABG was

shown, and admissions for stroke and PCI increased moderately.

There was a considerable male excess in both diabetic and non-diabetic groups for all

outcomes except for stroke (Table 1). The greatest male excess occurred in admissions for

cardiovascular interventions in both groups. The female-to-male ratio significantly decreased

for all outcomes during the study period representing an increasing male predominance. As

expected, most admissions for CVD events and procedures occurred among people over 65

years of age. Notably, women admitted for AMI were 5 years older on average in the diabetes

group and 9 years older in the non-diabetic group compared with men in corresponding

groups. Length of hospital stay (LOS) significantly reduced for AMI, stroke and PCI

admissions in the diabetes and non-diabetes groups in both sexesmen and women between

2004 and 2014. For CABG, while a reduction occurred in LOS in all groups, it remained

unchanged in women without diabetes.

Inpatient mortality rates

There was a statistically significant decline in inpatient mortality rates in people with and

without diabetes between 2004 and 2014 in all CVD outcomes except for PCI (Table 1).

11

Although the decline in inpatient mortality was evident in both men and women, women had

higher in-hospital mortality in both groups for all outcomes throughout the study period.

Changes in admission rates

Figure 1 shows the age- and sexgender-standardised rates for all study outcomes over the 11-

year study period in people with and without diabetes. Men had higher absolute rates in both

groups for all outcomes except for stroke in the non-diabetes group (Figure 1). Although age-

and sexgender- standardised admission rates for AMI showed reductions in the diabetes

group, these temporal changes did not reach statistical significance in regression models after

adjustment for study covariates (Figure 1, Table 2). Diabetes-related admission rates for

stroke and PCI increased significantly with rate ratios of 1.02 (95% CI 1.01-1.03) and 1.03

(10.01-1.04), representing an annual an increase of 2% and 3%, respectively (Figure 1, Table

2). In people without diabetes, admissions for AMI declined statistically significantly (IRR

(95% CI) 0.98 (0.95-0.99)) and remained stable for stroke and PCI. Both groups experienced

significant reductions in CABG rates.

During the eleven-year study period, people with and without diabetes experienced similar

proportional changes in admission rates for all study outcomes (Table 2). There were no

statistically significant differences in trends between women and men in either group leaving

the relative risk of events unchanged during the study period (Table 2).

Relative risk of CVD

Table 3 shows the Ratio Ratios for study outcomes obtained from negative binomial

regression models comparing people with and with diabetes by gender (using men as

reference) and comparing men and women by diabetes status (using people without diabetes

12

as reference). Among women without diabetes, admission rates were approximately three-

times lower for AMI, 14% lower for stroke, 3.7-times lower for PCI and five-times lower for

CABG compared with males without diabetes, after adjustment for study covariates (Table

3).

In adjusted models, women with diabetes had 44% higher rates of admission for AMI (IRR

95% CI, 1.44 (1.24-1.67)) and an approximately 2-fold higher rates of stroke (1.94 (1.67-

2.24)) compared with men without diabetes. Similarly, women with diabetes had significantly

higher rates of PCI and CABG compared with men without diabetes during the study period.

By contrast, to the 3-times lower admission rates for AMI among women compared with men

in the no in the presence of diabetes group, women with diabetes had approximately half of

the admission rates for AMI (IRR 95% CI 0.46 (0.40-0.53)) relative to men with diabetes.

Although women with diabetes were significantly less likely to undergo a PCI or CABG

compared with men with diabetes, the risk reduction is lower than that seen when comparing

men and women without diabetes (Table 3). These figures may represent a moderate

attenuation of the female advantage seen in the no diabetes group for admissions for AMI and

revascularisation procedures. Women with diabetes had 27% lower admission rates for stroke

and were statistically significantly less likely to undergo a PCI or CABG compared with men

with diabetes (Table 3).

While the presence of diabetes in men tripled the risk of being admitted for AMI (IRR (95%

CI) 3.15 (2.72-3.64)) and increased the risk of stroke admissions by 2.7-times (2.67 (2.31-

3.09)) (i.e. men with diabetes vs. men without diabetes), in women diabetes was associated

with a 4.3-fold increased risk of admission for AMI (IRR (95% CI) 4.273 (3.8-4.93),

13

P<0.001) and a 2.3-fold increased risk of stroke (2.329 (1.94-2.60), P<0.001) (i.e. women

with diabetes vs. women without diabetes).. Similarly, while in men diabetes was associated

with an approximately 3-fold increased rates of PCI and 5-fold increased rates of CABG,

among women diabetes was associated with 4-fold increased PCI admission rates and 6.2-

fold increased CABG rates (Table 3).

Inpatient mortality rates

There was a statistically significant decline in inpatient mortality rates in people with and

without diabetes between 2004 and 2014 for all study outcomes except for PCI (Tables 1 and

2). The results of multivariate negative binomial regression models show that inpatient

mortality rates are lower among women compared with men in both people with and without

diabetes for all outcomes but stroke admissions (Table 3). Men with diabetes experienced

lower rates of inpatient mortality related to AMI and stroke admissions than men without

diabetes after adjustment for study co-variates. By contrast, women with diabetes had higher

rates of inpatient mortality following AMI admissions compared with females free of

diabetes (Table 3).

14

Discussion

Our data revealed diverse trends in hospital admissions for major CVD events and coronary

revascularisation procedures in men and women with and without diabetes between 2004 and

2014 in England. Among people with diabetes, rates of hospital admissions for AMI

remained unchanged, increased for stroke and PCI and declined for CABG over the eleven-

year study period. By contrast, people without diabetes experienced reductions in admission

rates for AMI and CABG but not for stroke and PCI. However, these changes were modest

and tThere were no statistically significant differences in trends between people with and

without diabetes during the study period for any of the study outcomes except for CABG.

Our analyses of national data showed that although the absolute admission rates were higher

among men for all outcomes, gender protection among women did not disappear in diabetes,

it attenuated moderately for AMI but not for strokediabetes among women was associated

with a greater risk of hospital admissions for AMI and revascularisation procedures but not

for stroke. While the presence of diabetes was associated with tripled admission rates for

AMI in men, it increased rates by 4.3-fold among women. Similarly, while men with diabetes

had 3-times higher admission rates for PCI and 5-times higher admission rates for CABG,

among women the presence of diabetes was associated with 4.4-fold increased PCI and 6.2-

fold increased CABG rates compared with women without diabetes. Men and women

experienced similar proportional changes in admission rates for all study outcomes, leaving

the relative risk of events unchanged during the study period. Women with and without

diabetes experienced lower inpatient mortality rates than men for all admissions.

Temporal changes in cardiovascular morbidity related to diabetes

The past decade has seen immense advances in targeted and population-wide approaches to

the prevention and management of CVD (22). However, it is not known how these efforts

15

have affected contemporary patterns of CVD morbidity in England, and earlier studies may

not provide accurate reflections on current trends. Many developed countries have

documented downward trends in acute CVD events and related mortality in people with

diabetes (23-26). Some studies reported larger reductions in AMI and stroke rates in diabetes

and, therefore, narrowing differences in event rates between people with and without diabetes

(23, 24). Our results correspond with those studies that found similar temporal changes in

CVD event rates people with and without diabetes (20, 25).

Our observations of unchanged hospital admission rates of for AMI and increased stroke

rates in diabetes are important as there have been considerable investments in England

targeting the early detection and management of diabetes and its complications, and the

optimisation of CVD risk factor control. Although it is difficult to make comparisons across

studies due to differences in case definitions and observation periods, this finding differs

from the results of reduced diabetes-related admissions for AMI and unchanged stroke rates

found by our previous analyses over a shorter observation period, between 2004 and 2009

(20). The Quality and Outcomes Framework (QOF), the world’s largest pay-for-performance

scheme in primary care that was introduced in 2004 in England, resulted in improved

electronic recording, advanced processes of care and better intermediate outcomes for

diabetes (27). However, it has been contested whether these improvements were due to pre-

existing trends, and to what extent they translated into better clinical outcomes for patients

(such as prevention of AMI and stroke) (28). The recorded attenuation of improvements in

intermediate outcomes after the early years of QOF (including blood pressure, cholesterol

and glycated haemoglobin levels) may explain the levelling off of potential clinical benefits

over a longer period of time (27, 28). Furthermore, advances in clinical management may not

have detectable impact on hard clinical outcomes due to other important influences such as

16

changes in the makeup of people with diabetes, longer life expectancy of patients and the

socio-economically patterned distribution of comorbidity across population subgroups (29).

The recorded fall in mean blood pressure, cholesterol, glycated haemoglobin levels and

smoking prevalence may also be off-set by the increase in body mass index (30).

Increase in admissions among people with diabetes for stroke might reflect improving

survival of patients following cardiac events and longer life expectancy rather than the

insufficiency of stroke prevention (23). The finding of increased volume of coronary

revascularisations with downward trends in the use of CABG and a parallel increase in PCI

corresponds with previous studies (31, 32). These trends may reflect a substantial shift

towards less invasive revascularisation procedures (31).

Gender differences in the risk of cardiovascular morbidity associated with diabetes

Strong evidence suggests that dDiabetes confers a greater excess risk of cardiovascular

morbidity and mortality in women compared with men (6-8). While some, but not all, studies

reported equal or higher absolute risk of cardiovascular disease in women with diabetes

relative to men, studies have repeatedly documented a greater effect of diabetes on CVD

outcomes among female compared with male patients (7, 9, 11). The results of our study

correspond with those that observed a greater absolute risk of cardiovascular outcomes

among men compared with women with diabetes but found a higher relative risk of coronary

heart disease related to diabetes in women compared with men (7, 9, 33). While diabetes was

associated with tripled admission rates for AMI in men, it conferred an over 4-fold increase

in rates among women. Somewhat lower figures were reported by a meta-analysis that found

that the relative risk of incident CHD in people with diabetes relative to those without

diabetes was 2.16 in men and 2.82 in women (6). We are not aware of other nationalwide

17

studies investigating sexgender differences in hospital admissions for major CVD events

causes associated with diabetes over a longer period of time in England. The findings of our

national study differ from previous analyses that reported equalised or increased risk of AMI

among women with diabetes compared with men with diabetes . A cohort study from the UK

found a moderately higher risk of non-fatal AMI among women compared with men with

Type 2 diabetes but only among younger people below the age of 60 years (34).

However, we found that sex differences favouring women attenuated for admissions for AMI

in diabetes. While diabetes was associated with tripled admission rates for AMI in men, it

conferred an over 4-fold increase in rates among women. Somewhat lower figures were

reported by a meta-analysis that found that the relative risk of incident CHD in people with

diabetes relative to those without diabetes was 2.16 in men and 2.82 in women .

Women with diabetes in our study had slightly lower rates of stroke compared with men with

diabetes after adjustment for age, and the presence of diabetes was not associated with greater

admission rates for stroke in female compared with male patients. While diabetes conferred a

2.7-fold increased rates for stroke admissions in men, it increased rates by 2.3-fold among

women. The similar excess risk for stroke among women and men corresponds with the

results of some previous studies and has been suggested to The lack of sex differences in the

effect of diabetes on stroke may be associated withfollow from the reduced impact of

diabetes on CVD risk with an increasing age (5, 9, 34).

Although our study does not provide explanations for the stronger association between

diabetes and AMI among women relative to men, several potential mechanisms have been

suggested by previous studies (5, 35). These mechanisms include physiological, clinical and

management factors related to CVD that have been suggested to differ by gender. Many

18

studies observed a greater deterioration in a range of conventional and novel cardiovascular

risk factors in women compared with men while progressing from a non-diabetic to a diabetic

state (5, 11, 36). These risk factors encompass inflammation, impaired fibrinolysis,

coagulation, dyslipidaemia, hypertension and endothelial dysfunction, amongst other

metabolic and hormonal factors (11, 36-38). Gender differences in the association between

diabetes and CVD risk factors have been partly attributed to differences in central adiposity

and insulin resistance (11, 12, 35, 37) . Women tend to have peripheral fat distribution by

contrast to the central obesity with visceral fat accumulation found in men, resulting in

improved insulin sensitivity even at greater levels of weight gain (11). Women, therefore,

need to undergo more profound changes in metabolic risk factors (including body mass index

and insulin resistance) to develop diabetes and tend to present a more adverse CVD risk

factor profile at diagnosis (12, 36).

It has been widely documented that tThe heavier risk factor burden among women is

compounded by the suboptimal use of diagnostic procedures and evidence-based

interventions, less aggressive treatment strategies and a lower attainment of clinical treatment

targets during the course of diabetes (7, 35, 39). These gender disparities have been

repeatedly documented in England and worldwide and seem to persist over time (5, 15, 35,

40). There are sex differences in the way diabetes affects CVD risk throughout the course of

the disease. The worse risk factor profile among women with diabetes has been attributed to

the differential use of evidence-based preventive strategies and potential differences in the

impact of these preventive efforts on clinical outcomes . As an example, a large cohort study

from Women with diabetes in England were found that women with diabetesto have poorer

blood pressure, cholesterol and glycaemic control compared with men with diabetes between

1997 and 2005 . Women with diabetes were also less likely than men to have their glycated

19

haemoglobin and cholesterol recorded, and were less likely to be prescribed antiglycaemic

and lipid-lowering medications and had poorer blood pressure, cholesterol and glycaemic

control (15). Some of these gender disparities may stem from the notion that women are

‘protected’ against CVD resulting in the underestimation of the risk of serious acute CVD

events and long-term CVD risk, and may also affect women’s self-awareness of their own

risk (35). Gender Other contributory factors include that women tend to have differences in

the an ‘atypical’ clinical presentation of acute coronary syndrome may result in delayed or

missed diagnoses (35). Furthermore, (including the lack of chest pain) and some previous

reports suggesting some early reports suggesting that women may benefit less than men from

certain revascularisation procedures and pharmacological treatments in CVD risk reduction

that may have led to , may the underuse of preventative and therapeuticclinical interventions

(41).

Gender disparities in cardiovascular disease outcomes have important clinical and public

health implications. Despite the repeated documentations of a more adverse impact of

diabetes on CVD profile and higher relative risk of fatal and non-fatal CHD among women,

CVD is still under-recognised and undertreated among women (35, 41). Importantly, women

with AMI have been reported to have more unfavourable outcomes following an AMI

including higher in-hospital and 1one-year mortality (35). Although not specifically assessed

by our study, previous analyses highlighted a greater relative risk of AMI and higher

complication and mortality rates following AMI, specifically among younger women

compared with similarly aged men (35, 41). These acknowledged these couldgender

disparities could serve as potential targets for interventions. However, clinical guidelines and

quality improvement initiatives do not specifically recognise the unique risk factor profile of

women with diabetes (16). The ‘Action for Diabetes’ a document outlining NHS England’s

20

vision for future diabetes care, did not consider potential sex differences in the effect of

diabetes Recognising this gap, e clinical and public health significance of sex differences, the

American Heart Association published a scientific statement and recommendations for earlier

and more targeted interventions for the prevention of CVD among women (42).

Strengths and limitations

To our knowledge, this is the first national study describing recent trends of hospital

admissions for CVD outcomes in men and women by diabetes status England over the past

decade. Our study covers the entire population of England and given that all study outcomes

require hospital admission, our results are likely to provide an accurate reflection on the

burden of CVD in people with diabetes. Furthermore, we had data on the number of people

with diabetes in England for each study year, allowing the estimation of diabetes-specific

event rates.

Limitations of our study include that we cannot directly evaluate miscoding, misdiagnosis

and misclassification of diagnoses in the national HES database. HES is a financial database,

and However, given that the reimbursement of NHS hospitals for each patient seen is directly

determined by coding data, hospitals have a strong financial incentive to establish and

maintain accurate coding practices. rRoutinely collected data are subject to regular national

clinical coding audits continuously audited and a systematic review of discharge coding

evaluated its accuracy high for both diagnoses and procedures (43). HES do not include data

on the clinical, laboratory and lifestyle characteristics of individuals admitted to NHS

hospitals in England. Therefore, we were unable to adjust our analyses for changes in

cardiovascular risk factors of patients requiring hospital admission for CVD causes during the

study period. However, analysing national hospital activity data in consecutive years provides

21

an important and accurate reflection on temporal changes in the burden of major CVD events

and procedures at a national level. We did not report Type 1 and Type 2 diabetes-specific

events separately because national diabetes registers do not include information on diabetes

type. We used data from Health Survey for England to ascertain the age and sexgender

distribution of diabetes in England. For 2004, 2005 and 2007 and 2008 when age- and

sexgender-specific data were not available we used data from 2006, the mid-term of this 5-

year period.

Conclusions

In conclusion, in this national study covering the entire population of England, we found

unchanged admission rates for AMI and increased rates for stroke in people with diabetes

between 2004 and 2014. The consistent upward trend in the number of diabetes-related

admissions reflect the increasing absolute burden of CVD complications in diabetes.

Although women with diabetes had lower absolute CVD event rates compared with men,

diabetes conferred a larger relative increase in risk for hospital admissions for AMI in women

compared with men. These observations call for aligning CVD risk with the intensity of

preventive strategies in England and intensified a more aggressive cardiovascular risk factor

management among women.

Declarations

Ethics approval and consent to participate

We have approval from the Secretary of State and the Health Research Authority under

Regulation 5 of the Health Service (Control of Patient Information) Regulations 2002 to hold

confidential data and analyse them for research purposes (CAG ref 15/CAG/0005). We have

22

approval to use them for research and measuring quality of delivery of healthcare, from the

London - South East Ethics Committee (REC ref 15/LO/0824).

Consent for publication

Not applicable.

Availability of data and materials

Publicly available data sources used:

Mid-year population estimates for England for each study year are available from the Office

for National Statistics at:

https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/

populationestimates/datasets/

populationestimatesforukenglandandwalesscotlandandnorthernireland

Data on the number of people with diagnosed diabetes in England from the Quality and

Outcomes Framework are available at: http://content.digital.nhs.uk/qof

Non-publicly available data sources:

Other data sources used in this study (Hospital Episode Statistics and Health Survey for

England) are not publicly available. Under our user agreement with the data provider we

cannot release patient data from these data sources.

Declaration of conflict of interests

23

None.

Funding statement

CM is funded by an NIHR Research Professorship.

Authors’ contributions

AL wrote the manuscript and researched the data. AB reviewed/edited the manuscript and

researched the data. SK researched the data and edited the manuscript. BV reviewed/edited

the manuscript and contributed to the discussion. AM reviewed/edited the manuscript and

revised it for important intellectual content. CM wrote the manuscript and revised it for

important intellectual content. EV conceived the idea, wrote the manuscript and researched

the data. All authors gave final approval of this version to be published.

Acknowledgements

The Department of Primary Care & Public Health at Imperial College is grateful for support

from the NIHR North West London CLAHRC, the NIHR Biomedical Research Centre, and

the Imperial Centre for Patient Safety and Service Quality.

24

References

1. National Diabetes Audit. Report 2: Complications and Mortality, 2015 Available at:

http://content.digital.nhs.uk/catalogue/PUB16496/nati-diab-audi-12-13-rep2.pdf (Last

accessed: 3rd July 2017)

2. van Dieren S, Beulens JW, van der Schouw YT, Grobbee DE, Neal B. The global

burden of diabetes and its complications: an emerging pandemic. European journal of

cardiovascular prevention and rehabilitation : official journal of the European Society of

Cardiology, Working Groups on Epidemiology & Prevention and Cardiac Rehabilitation and

Exercise Physiology. 2010;17 Suppl 1:S3-8.

3. Kappert K, Bohm M, Schmieder R, Schumacher H, Teo K, Yusuf S, et al. Impact of

sex on cardiovascular outcome in patients at high cardiovascular risk: analysis of the

Telmisartan Randomized Assessment Study in ACE-Intolerant Subjects With Cardiovascular

Disease (TRANSCEND) and the Ongoing Telmisartan Alone and in Combination With

Ramipril Global End Point Trial (ONTARGET). Circulation. 2012;126(8):934-41.

4. Anand SS, Islam S, Rosengren A, Franzosi MG, Steyn K, Yusufali AH, et al. Risk

factors for myocardial infarction in women and men: insights from the INTERHEART study.

European heart journal. 2008;29(7):932-40.

5. Peters SA, Huxley RR, Sattar N, Woodward M. Sex Differences in the Excess Risk of

Cardiovascular Diseases Associated with Type 2 Diabetes: Potential Explanations and

Clinical Implications. Current cardiovascular risk reports. 2015;9(7):36.

6. Peters SA, Huxley RR, Woodward M. Diabetes as risk factor for incident coronary

heart disease in women compared with men: a systematic review and meta-analysis of 64

cohorts including 858,507 individuals and 28,203 coronary events. Diabetologia.

2014;57(8):1542-51.

25

7. Huxley R, Barzi F, Woodward M. Excess risk of fatal coronary heart disease

associated with diabetes in men and women: meta-analysis of 37 prospective cohort studies.

Bmj. 2006;332(7533):73-8.

8. Peters SA, Huxley RR, Woodward M. Diabetes as a risk factor for stroke in women

compared with men: a systematic review and meta-analysis of 64 cohorts, including 775,385

individuals and 12,539 strokes. Lancet. 2014;383(9933):1973-80.

9. Ballotari P, Venturelli F, Greci M, Giorgi Rossi P, Manicardi V. Sex Differences in

the Effect of Type 2 Diabetes on Major Cardiovascular Diseases: Results from a Population-

Based Study in Italy. Int J Endocrinol. 2017;2017:6039356.

10. Anagnostis P, Majeed A, Johnston DG, Godsland IF. Cardiovascular risk in women

with type 2 diabetes mellitus and prediabetes: is it indeed higher than men? European journal

of endocrinology. 2014;171(6):R245-55.

11. Wannamethee SG, Papacosta O, Lawlor DA, Whincup PH, Lowe GD, Ebrahim S, et

al. Do women exhibit greater differences in established and novel risk factors between

diabetes and non-diabetes than men? The British Regional Heart Study and British Women's

Heart Health Study. Diabetologia. 2012;55(1):80-7.

12. Logue J, Walker JJ, Colhoun HM, Leese GP, Lindsay RS, McKnight JA, et al. Do

men develop type 2 diabetes at lower body mass indices than women? Diabetologia.

2011;54(12):3003-6.

13. Bugiardini R, Yan AT, Yan RT, Fitchett D, Langer A, Manfrini O, et al. Factors

influencing underutilization of evidence-based therapies in women. European heart journal.

2011;32(11):1337-44.

14. Kramer HU, Raum E, Ruter G, Schottker B, Rothenbacher D, Rosemann T, et al.

Gender disparities in diabetes and coronary heart disease medication among patients with

type 2 diabetes: results from the DIANA study. Cardiovascular diabetology. 2012;11:88.

26

15. Hamilton FL, Bottle A, Vamos EP, Curcin V, Ng A, Molokhia M, et al. Impact of a

pay-for-performance incentive scheme on age, sex, and socioeconomic disparities in diabetes

management in UK primary care. The Journal of ambulatory care management.

2010;33(4):336-49.

16. National Institute for Health and Care Excellence: Type 2 diabetes in adults:

management. Available at: https://www.nice.org.uk/guidance/ng28 (Last accessed: 3rd July

2017).

17. National Health Service Digital: Quality and Outcomes Framework. Available at:

http://content.digital.nhs.uk/qof (Last accessed: 3rd July 2017).

18. Health Survey for England. Available at:

https://data.gov.uk/dataset/health_survey_for_england (Last accessed: 3rd July 2017).

19. Vamos EP, Bottle A, Edmonds ME, Valabhji J, Majeed A, Millett C. Changes in the

incidence of lower extremity amputations in individuals with and without diabetes in England

between 2004 and 2008. Diabetes care. 2010;33(12):2592-7.

20. Vamos EP, Millett C, Parsons C, Aylin P, Majeed A, Bottle A. Nationwide study on

trends in hospital admissions for major cardiovascular events and procedures among people

with and without diabetes in England, 2004-2009. Diabetes care. 2012;35(2):265-72.

21. Office for National Statistics: Population Estimates for UK, England and Wales,

Scotland and Northern Ireland Available at:

https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/

populationestimates/datasets/

populationestimatesforukenglandandwalesscotlandandnorthernireland (Last accessed: 3rd

July 2017).

27

22. Mortality GBD, Causes of Death C. Global, regional, and national age-sex specific

all-cause and cause-specific mortality for 240 causes of death, 1990-2013: a systematic

analysis for the Global Burden of Disease Study 2013. Lancet. 2015;385(9963):117-71.

23. Gregg EW, Li Y, Wang J, Burrows NR, Ali MK, Rolka D, et al. Changes in diabetes-

related complications in the United States, 1990-2010. The New England journal of

medicine. 2014;370(16):1514-23.

24. Booth GL, Kapral MK, Fung K, Tu JV. Recent trends in cardiovascular complications

among men and women with and without diabetes. Diabetes care. 2006;29(1):32-7.

25. Fox CS, Coady S, Sorlie PD, Levy D, Meigs JB, D'Agostino RB, Sr., et al. Trends in

cardiovascular complications of diabetes. Jama. 2004;292(20):2495-9.

26. Jansson SP, Andersson DK, Svardsudd K. Mortality trends in subjects with and

without diabetes during 33 years of follow-up. Diabetes care. 2010;33(3):551-6.

27. Calvert M, Shankar A, McManus RJ, Lester H, Freemantle N. Effect of the quality

and outcomes framework on diabetes care in the United Kingdom: retrospective cohort study.

Bmj. 2009;338:b1870.

28. Doran T, Kontopantelis E, Valderas JM, Campbell S, Roland M, Salisbury C, et al.

Effect of financial incentives on incentivised and non-incentivised clinical activities:

longitudinal analysis of data from the UK Quality and Outcomes Framework. Bmj.

2011;342:d3590.

29. Gregg EW, Sattar N, Ali MK. The changing face of diabetes complications. The

lancet Diabetes & endocrinology. 2016;4(6):537-47.

30. Unal B, Critchley JA, Capewell S. Explaining the decline in coronary heart disease

mortality in England and Wales between 1981 and 2000. Circulation. 2004;109(9):1101-7.

28

31. Culler SD, Kugelmass AD, Brown PP, Reynolds MR, Simon AW. Trends in coronary

revascularization procedures among Medicare beneficiaries between 2008 and 2012.

Circulation. 2015;131(4):362-70; discussion 70.

32. Hassan A, Newman A, Ko DT, Rinfret S, Hirsch G, Ghali WA, et al. Increasing rates

of angioplasty versus bypass surgery in Canada, 1994-2005. American heart journal.

2010;160(5):958-65.

33. Juutilainen A, Kortelainen S, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Gender

difference in the impact of type 2 diabetes on coronary heart disease risk. Diabetes care.

2004;27(12):2898-904.

34. Shah AD, Langenberg C, Rapsomaniki E, Denaxas S, Pujades-Rodriguez M, Gale

CP, et al. Type 2 diabetes and incidence of cardiovascular diseases: a cohort study in 1.9

million people. The lancet Diabetes & endocrinology. 2015;3(2):105-13.

35. Maas AH, Appelman YE. Gender differences in coronary heart disease. Neth Heart J.

2010;18(12):598-602.

36. Gomez-Marcos MA, Recio-Rodriguez JI, Gomez-Sanchez L, Agudo-Conde C,

Rodriguez-Sanchez E, Maderuelo-Fernandez J, et al. Gender differences in the progression of

target organ damage in patients with increased insulin resistance: the LOD-DIABETES

study. Cardiovascular diabetology. 2015;14:132.

37. Buday B, Pach PF, Literati-Nagy B, Vitai M, Kovacs G, Vecsei Z, et al. Sex

influenced association of directly measured insulin sensitivity and serum transaminase levels:

Why alanine aminotransferase only predicts cardiovascular risk in men? Cardiovascular

diabetology. 2015;14:55.

38. Sung KC, Chang Y, Ryu S, Chung HK. High levels of serum vitamin D are associated

with a decreased risk of metabolic diseases in both men and women, but an increased risk for

coronary artery calcification in Korean men. Cardiovascular diabetology. 2016;15(1):112.

29

39. Ford ES, Ajani UA, Croft JB, Critchley JA, Labarthe DR, Kottke TE, et al.

Explaining the decrease in U.S. deaths from coronary disease, 1980-2000. The New England

journal of medicine. 2007;356(23):2388-98.

40. Regensteiner JG, Golden S, Huebschmann AG, Barrett-Connor E, Chang AY, Chyun

D, et al. Sex Differences in the Cardiovascular Consequences of Diabetes Mellitus: A

Scientific Statement From the American Heart Association. Circulation. 2015;132(25):2424-

47.

41. Khamis RY, Ammari T, Mikhail GW. Gender differences in coronary heart disease.

Heart. 2016;102(14):1142-9.

42. Mosca L, Benjamin EJ, Berra K, Bezanson JL, Dolor RJ, Lloyd-Jones DM, et al.

Effectiveness-based guidelines for the prevention of cardiovascular disease in women--2011

update: a guideline from the American Heart Association. J Am Coll Cardiol.

2011;57(12):1404-23.

43. Burns EM, Rigby E, Mamidanna R, Bottle A, Aylin P, Ziprin P, et al. Systematic

review of discharge coding accuracy. Journal of public health. 2012;34(1):138-48.

Figure legends

Figure 1. Age- and sexgender-standardized rates of admissions by year and sexgender

A. Acute myocardial infarction; B. Stroke; C. Percutaneous coronary intervention; D. Coronary artery

bypass grafting. Rates are expressed as 100,000 people with diabetes (Diabetes group) or 100,000

people without diabetes (Non diabetes group)

30

Table 1. Characteristics of people affected by cardiovascular disease by diabetes status and sexgender in England between 2004-2005 and 2014-2015

Diabetes No diabetes

Women Men Women Men

Outcome 2004-2005 2014-2015 P 2004-2005 2014-2015 P 2004-2005 2014-2015 P 2004-2005 2014-2015 P

AMI

Number of events 4,848 5,982 7,641 10,843 22,811 18,735 38,305 35,083

% of males - - 61.2 64.5 <0.001a - - 62.7 65.2 <0.001a

Female-to-male ratio (95% CI)

Age, yr, mean (SD)

-

74.9 (11.1)

-

73.7 (12.7) <0.001b

0.63 (0.62-0.64)

69.3 (11.8)

0.55 (0.54-0.56)

69.4 (12.5)

<0.001a

0.7273b

-

76.2 (12.3)

-

75.2 (14.0) 0.001b

0.59 (0.59-0.6)

67.1 (13.7)

0.53 (0.53-0.54)

66.3 (14.1)

<0.001a

0.001b

Age groups, N (%)

17-44 years 72 (1.5) 121 (2.0) <0.001a 211 (2.8) 292 (2.7) <0.001a 406 (1.8) 437 (2.3) <0.001a 2,198 (5.7) 1,986 (5.7) <0.001a

45-64 years 709 (14.6) 1,270 (21.2) 2,203 (28.8) 3,434 (31.7) 3,448 (15.1) 3,854 (20.6) 13,726 (35.8) 14,127 (40.3)

65 years 4,067 (83.9) 4,591 (76.8) 5,227 (68.4) 7,117 (65.6) 18,957 (83.1) 14,444 (77.1) 22,381 (58.5) 18,970 (54.0)

Median LOS, days (IQR)

5 (8-14) 3 (5-10) <0.001d 5 (7-12) 3 (4-8) <0.001d 5 (7-12) 2 (4-8) <0.001d 4 (6-10) 2 (3-6) <0.001d

Number of deaths (%) 881 (18.2) 676 (11.3) <0.001a 1,031 (13.5) 954 (8.8) <0.001a 4,417 (19.4) 1,982 (10.6) <0.001a 4,453 (11.6) 2,347 (6.7) <0.001a

Stroke

Number of events 4,702 7,678 4,959 9,600 34,079 34,947 29,005 33,247

% of males - - 51.3 55.6 <0.001a - - 46.0 48.7 <0.001a

Female-to-male ratio (95% CI)

Age, yr, mean (SD)

-

76.2 (11.1)

-

76.8 (11.9) 0.006b

0.95 (0.94-0.95)

72.0 (11.1)

0.80 (0.80-0.81)

72. 5 (11.6)

<0.001a

0.002b

-

77.1 (13. 4)

-

76.5 (14. 6) <0.001b

1.17 (1.17-1.2)

71.4 (13.6)

1.05 (1.04-1.05)

71.1 (14.4)

<0.001a

0.002b

31

Age groups, N (%)

17-44 years 58 (1.2) 99 (1.3) 0.0436a 82 (1.6) 176 (1.8) 0.405a 1,116 (3.3) 1,288 (3.7) <0.001a 1,353 (4.7) 1,628 (4.9) <0.001a

45-64 years 589 (12.5) 1,086 (14.1) 1,038 (20.9) 2,081 (21.7) 4,147 (12.2) 5,181 (14.8) 6,478 (22.3) 8,151 (24.5)

65 years 4,055 (86.2) 6,493 (84.6) 3,839 (77.4) 7,343 (76.5) 28,816 (84.6) 28,478 (81.5) 21,174 (73.0) 23,468 70.6)

Median LOS, days (IQR)

6 (15-33) 3 (7-20) <0.001d 5 (11-28) 2 (6-16) <0.001ε 5 (12-29) 2 (7-19) <0.001d 4 (10-24) 2 (5-14) <0.001d

Number of deaths (%) 1,485 (31.6) 1,438 (18.7) <0.001a 1,191 (24.0) 1,410 (14.7) <0.001a 11,614 (34.1) 7,005 (20.4) <0.001a 7,476 (19.5) 4,812 (13.7) <0.001a

CABG

Number of events 1,028 1,111 3,531 4,275 4,029 2,649 15,165 11,005

% of males - - 77.5 79.4 0.02a - - 79.0 80.6 <0.001a

Female-to-male ratio

(95% CI)

Age, yr mean (SD)

-

67.5 (8.5)

-

68.1 (9.8) 0.105b

0.29 (0.27-0.3)

65.1 (8.7)

0.26 (0.25-0.27)

66.7 (9.5)

0.02a

<0.001b

-

69.2 (9.1)

-

70.3 (10.1) <0.001b

0.26 (0.25-0.27)

65.4 (9.4)

0.24 (0.23-0.25)

67.1 (10.2)

<0.001a

<0.001b

Age groups, N (%)

17-44 years 13 (1.3) 16 (1.4) 0.940a 57 (1.6) 64 (1.5) 0.004a 63 (1.6) 40 (1.5) 0.610a 317 (2.1) 199 (1.8) <0.001a

45-64 years 322 (31.3) 348 (31.3) 1,460 (41.3) 1,612 (37.7) 966 (24.0) 608 (30.0) 6,217 (41.0) 3,916 (35.6)

65 years 693 (67.4) 747 (67.2) 2,014 (57.0) 2,599 (60.8) 3,000 (74.5) 2,001 (75.5) 8,631 (56.9) 6,890 (62.6)

Median LOS, days (IQR)

8 (11-20) 8 (11-18) 0.086d 7 (9-16) 7 (9-14) <0.001d 7 (10-16) 7 (10-17) 0.672 c 7 (8-13) 6 (9-14) 0.041 c

Number of deaths (%) 53 (5.1) 30 (2.7) 0.003a 102 (2.8) 69 (1.6) <0.001a 219 (5.4) 131 (4.9) 0.3479a 412 (2.7) 256 (2.3) 0.054a

PCI

Number of events 2,238 4,576 5,106 11,862 11,190 14,764 31,386 43,178

% of males - - 69.5 72.2 <0.001a - - 73.7 74.5 0.004a

Female-to-male ratio

(95% CI)

- - 0.44 (0.42-0.45) 0.38 (0.37-0.39) <0.001a - - 0.36 (0.35-0.37) 0.34 (0.33-0.34) 0.004a

32

Age, yr, mean (SD) 65.3 (10.7) 67.9 (11.8) <0.001b 63.2 (10.1) 65.5 (11.2) <0.001b 66.4 (10.5) 68.9 (12.2) <0.001b 61.7 (10.9) 63.4 (12.2) <0.001b

Age groups, N (%)

17-44 years 91 (4.1) 152 (3.3) <0.001a 217 (4.2) 402 (3.4) <0.001a 348 (3.1) 416 (2.8) <0.001a 1,996 (6.4) 2,438 (5.6) <0.001a

45-64 years 850 (38.0) 1,513 (33.1) 2,404 (47.1) 4,982 (42.0) 4,066 (36.3) 4,585 (31.1) 16,445 (52.4) 20,416 (47.2)

65 years 1,297 (58.0) 2,911 (63.6) 2,485 (48.7) 6,478 (54.6) 6,776 (60.6) 9,763 (66.1) 12,945 (41.2) 20,324 (47.1)

Median LOS, days (IQR)

1 (2-6) 1 (2-5) 0.013d 1 (2-5) 1 (2-4) <0.001d 1 (2-4) 1 (2-4) <0.001d 1 (2-4) 1 (2-3) <0.001d

Number of deaths (%) 45 (2.0) 146 (3.2) 0.006a 65 (1.3) 253 (2.1) <0.001a 142 (1.3) 426 (2.9) <0.001a 282 (0.9) 851 (2.0) <0.001a

Note: AMI: acute myocardial infarction, CABG: Coronary artery bypass grafting; PCI: Percutaneous coronary intervention; LOS: Length of Hospital Stay; IQR: interquartile range; CI: confidence interval, SD: Standard Deviation.a Chi-square test; b Student’s t test; c Wilcoxon test – 2004 figure lower; d Wilcoxon test – 2004 figure higher

33

Table 2: Rate ratio of hospital admissions for CVD events and interventions in people with and without diabetes between 2004 and 2014 in England.

Changes in Rates

Total Male FemaleOutcome Rate ratio (95% CI)a P value P value for

diabetes vs. yearb

Rate ratio (95% CI)a P value Rate ratio (95% CI)a P value P value forsex vs. yearc

AMI Diabetes 0.99 (0.98-1.01) 0.59 0.31 1.00 (0.98-1.01) 0.63 0.99 (0.98-1.00) 0.10 0.47 No diabetes 0.97 (0.95-0.99) 0.04 0.97 (0.96-0.98) <0.001 0.98 (0.97-0.99) <0.001 0.99Stroke Diabetes 1.02 (1.01-1.03) 0.005 0.30 1.03 (1.01-1.04) <0.001 1.01 (1.00-1.02) 0.062 0.18 No diabetes 1.00 (0.99-1.01) 0.993 1.00 (0.99-1.01) 0.93 0.99 (0.99-1.01) 0.79 0.34PCI Diabetes 1.03 (1.01-1.04) <0.001 0.11 1.03 (1.02-1.04) <0.001 1.02 (1.01-1.03) 0.002 0.20 No diabetes 1.01 (0.97-1.03) 0.586 1.01 (1.00-1.02) 0.131 1.00 (0.99-1.01) 0.497 0.91CABG Diabetes 0.97 (0.96-0.98) <0.001 0.008 0.97 (0.96-0.98) <0.001 0.96 (0.95-0.97) <0.001 0.55 No diabetes 0.94 (0.93-0.95) <0.001 0.95 (0.94-0.95) <0.001 0.94 (0.93-0.95) <0.001 0.72

Changes in Rates

Hospital admissionsRate Ratio (95% CI)

In-patient mortalityRate Ratio (95% CI)

Total population Male Female Total populationOutcome Rate ratioa P P value

diabetes vs. yearb

Rate ratioa P Rate ratioa P P valuegender

vs. yearc

Rate ratioa P

AMI

34

Diabetes 0.99 (0.98-1.01) 0.587 0.311 1.00 (0.98-1.01) 0.626 0.99 (0.98-1.00) 0.096 0.470 0.95 (0.93-0.97) <0.001 No diabetes 0.97 (0.95-0.99) 0.039 0.97 (0.96-0.98) <0.001 0.98 (0.97-0.99) <0.001 0.992 0.96 (0.94-0.98) <0.001Stroke Diabetes 1.02 (1.01-1.03) 0.005 0.309 1.03 (1.01-1.04) <0.001 1.01 (1.00-1.02) 0.062 0.178 0.95 (0.93-0.96) <0.001 No diabetes 1.00 (0.99-1.01) 0.993 1.00 (0.99-1.01) 0.930 0.99 (0.99-1.01) 0.791 0.340 0.94 (0.93-0.96) <0.001PCI Diabetes 1.03 (1.01-1.04) <0.001 0.109 1.03 (1.02-1.04) <0.001 1.02 (1.01-1.03) 0.002 0.200 1.07 (1.05-1.09) <0.001 No diabetes 1.01 (0.97-1.03) 0.586 1.01 (1.00-1.02) 0.131 1.00 (0.99-1.01) 0.497 0.914 1.10 (1.07-1.12) <0.001CABG Diabetes 0.97 (0.96-0.98) <0.001 0.008 0.97 (0.96-0.98) <0.001 0.96 (0.95-0.97) <0.001 0.552 0.96 (0.94-0.98) <0.001 No diabetes 0.94 (0.93-0.95) <0.001 0.95 (0.94-0.95) <0.001 0.94 (0.93-0.95) <0.001 0.719 0.98 (0.97-0.99) <0.001

Note: AMI: acute myocardial infarction, CABG: Coronary artery bypass grafting; PCI: Percutaneous coronary intervention; IQR: interquartile range;

CI: confidence interval.a Rate ratios are obtained from negative binomial regression models adjusted for age, sexgender and study year. b P value is based on interaction between diabetes status and study year from negative binomial regression modelsc P value is based on interaction between sexgender and study year from negative binomial regression models

35

Table 3: Risk of major cardiovascular events and procedures in males and females with and without diabetes between 2004-2005 and 2014-2015 in England.

Rate Ratio (95% CI)

No diabetes a Diabetes b Male a Female c

Male(reference)

Female Male(reference)

Female No diabetes(reference)

Diabetes No diabetes(reference)

Diabetes

Admissions

AMI 1.00 0.33 (0.29-0.39) d 1.00 0.46 (0.40-0.53) d 1.00 3.15 (2.72-3.64) d 1.00 4.27 (3.78-4.93) d

Stroke 1.00 0.86 (0.75-0.99) e 1.00 0.73 (0.63-0.84) d 1.00 2.67 (2.31-3.09) d 1.00 2.29 (1.94-2.60) d

PCI 1.00 0.27 (0.24-0.30) d 1.00 0.37 (0.34- 0.42) d 1.00 3.14 (2.83-3.48) d 1.00 4.37 (3.93-4.85) d

CABG 1.00 0.22 (0.20-0.24) d 1.00 0.28 (0.25-0.30) d 1.00 5.01 (4.59-5.05) d 1.00 6.24 (5.66-6.88) d

In-patient mortality

AMI 1.00 0.59 (0.51-0.68) d 1.00 0.59 (0.51-0.68) d 1.00 0.68 (0.59-0.79) d 1.00 1.17 (1.01-1.34) f

Stroke 1.00 0.99 (0.86-1.15) 1.00 0.89 (0.76-1.04) 1.00 0.69 (0.59-0.80) d 1.00 0.62 (0.52-0.71) d

PCI 1.00 0.47 (0.41-0.52) d 1.00 0.52 (0.45- 0.60) d 1.00 1.18 (1.04-1.33) e 1.00 1.32 (1.15-1.51) d

CABG 1.00 0.48 (0.43-0.51) d 1.00 0.51 (0.45-0.57) d 1.00 0.93 (0.85-1.02) 1.00 1.01 (0.91-1.12)

Rate Ratio (95% CI)

No DM Male No DM Femalea Diabetes Malea Diabetes Femalea Diabetes Female

36

(reference) vs. Diabetes Male (reference)b

AMI 1.00 0.33 (0.29-0.39) c 3.15 (2.72-3.64) c 1.44 (1.24-1.67) c 0.46 (0.40-0.53) c

Stroke 1.00 0.86 (0.75-0.99) e 2.67 (2.31-3.09) c 1.94 (1.67- 2.24) c 0.73 (0.63-0.84) c

PCI 1.00 0.27 (0.24-0.30) c 3.14 (2.83-3.48) c 1.18 (1.06-1.31) d 0.37 (0.34- 0.42) c

CABG 1.00 0.22 (0.20-0.24) c 5.01 (4.59-5.05) c 1.39 (1.26-1.53) c 0.28 (0.25-0.30) c

Note: AMI: acute myocardial infarction, CABG: Coronary artery bypass grafting; PCI: Percutaneous coronary intervention; CI: confidence interval.a Rate ratios are obtained from negative binomial regression models adjusted for age, sex, diabetes status and study year with men without diabetes as reference. b Rate ratios for women vs. men with diabetes (reference) obtained from negative binomial regression models adjusted for age, sex, diabetes status and study year c P<0.001 d P<0.01 eP<0.05

Note: AMI: acute myocardial infarction, CABG: Coronary artery bypass grafting; PCI: Percutaneous coronary intervention; CI: confidence interval.a Rate ratios are obtained from negative binomial regression models adjusted for age, sex, diabetes status and study year with men without diabetes as reference. b Rate ratios are obtained from negative binomial regression models adjusted for age, sex, diabetes status and study year with men with diabetes as reference.c Rate ratios are obtained from negative binomial regression models adjusted for age, sex, diabetes status and study year with women without diabetes as reference.d P<0.001 e P<0.01 f P<0.05

37